ITS ACTION AT THE SURFACE. 329 
Mr. Bennett, floating at the surface, with the 
Upper portion of the shell raised above the water, 
und kept in a vertical position by means of the 
included air (see PI. 31. Fig. 1.); this position 
Js best adapted to the retrograde motion, which 
u Sepia derives from the violent ejection of 
Ulster through its funnel (k) ; thus far, the air- 
nhamhers, serve to maintain both the shell and 
body of the animal in a state of equilibrium al 
the surface. 
Secondly, The next point to be considered is 
the mode of operation of the siphuncle and air- 
chambers, in the act of sinking suddenly from 
the surface to the bottom. These are explained 
*n the note subjoined.* 
* It appears from the figure of the animal, PI. 34, with which 
t have been favoured by Mr. Owen, that the upper extremity * 
of the siphuncle marked by the insertion of the probe b., termi- 
nates in the cavity of the pericardium p, p. As this cavity may 
contain a fluid, more dense than water, excreted by the glandular 
follicles d. d., and is apparently of such a size that its contents 
tvould suffice to fill the siphuncle, it is probable that this fluid 
forms the circulating medium of adjustment, and regulates the 
ascent or descent of the animal by its interchange of place from 
the pericardium to the siphuncle. 
When the arms and body are expanded, the fluid remains in 
the pericardium, and the siphuncle is empty, and collapsed, and 
surrounded by the portions of air that are permanently confined 
"'ithin each air-chamber; in this state, the specific gravity of the 
body and shell together is such as to cause the animal to rise, 
und be sustained floating at the surface. 
When, on any alarm, the arms and body are contracted, and 
'ffawn into the shell, the retraction of these parts, causing pres- 
sure On the exterior of the pericardium, forces its fluid contents 
ownwards into the siphuncle ; and the bulk of the body being 
